hadamard-2/leyu-amharic-gojjam-dialect
Leyu Amharic - Gojjam Dialect Speech Corpus Dataset Description A parallel speech corpus of audio recordings paired with their transcripts, focused on the Gojjam dialect of Amharic, for ASR and TTS research. Leyu reports that recordings were collected from contributors on mobile devices in real-world environments, and that each audio–text pair was manually reviewed for transcript accuracy and audio clarity. This repository is a copy of… See the full description on the dataset page: https://huggingface.co/datasets/hadamard-2/leyu-amharic-gojjam-dialect.
docs: document the published split.json in the card
data: add leakage-free train/dev/test split assignment
docs: correct card figures to match the published parquet
Duplicate from gheero-Leyu/leyu-amharic-gojjam-dialect